Are you interested in exploring collaboration possibilities?
Enter your information in the form and a representative will contact you shortly.
Identifying Patients with Mild-Stage Emphysema in COPD via Machine Learning
May 17, 2025
Identifying biomarkers of mild-stage emphysema in COPD patients via interpretable machine learning
Introduction
Emphysema results from alveolar damage—causing abnormal extracellular matrix (ECM) remodeling and impaired lung function. Early detection in mild, often asymptomatic stages is key to timely intervention. Although computed tomography (CT) scans are the most accurate detection method, pathological changes may occur before emphysema becomes visible, highlighting the need for non-invasive early detection approaches.
This study aimed to develop a proof-of-concept machine learning (ML) pipeline to identify patients with mild-stage emphysema using circulating biomarkers.
Poster
Conclusion
ML demonstrates potential for early-stage emphysema diagnosis through
biomarker-driven methods.
Quantifying fragments of ECM remodeling—driven by immune cell activity and collagen formation—could potentially serve as early diagnostic biomarkers in patients without lung function decline.
